Proximal is a research lab focused on data.
Proximal is a new data company. Our core belief is that data which is complex enough to teach today’s frontier models is not bottlenecked by domain experts, but by research, great systems, and excellent software. We are not a recruiting firm or a talent marketplace, but a research and engineering organization that treats data as a problem which deserves the same level of rigor as work on training algorithms and model architectures.
We think that this is the most impactful work towards agents that can autonomously solve complex technical problems, and intend to share our research and progress in the open.
Announcements, benchmark releases, and the work behind both.
Research & Engineering•2 min read
FrontierSWE v2
34 tasks, 13 frontier models, 20-hour budgets. Our hardest coding benchmark yet.
| # | Model | Score |
|---|---|---|
| 1 | Claude Fable 5.156.3% | 56.3% |
| 2 | GPT-5.632.2% | 32.2% |
| 3 | GLM-5.330.2% | 30.2% |
| 4 | Kimi K325.9% | 25.9% |
| 5 | Grok 4.625.3% | 25.3% |
| 6 | Gemini 3.7 Flash20.3% | 20.3% |
| 7 | Qwen3.8-Max15.8% | 15.8% |
| 8 | DeepSeek V4 Flash Exp14.8% | 14.8% |
| 9 | Muse Spark 1.212.0% | 12.0% |
| 10 | Inkling4.1% | 4.1% |
Research & Engineering•9 min read
FrontierSWE
Our ultra-long-horizon coding benchmark. Frontier models clear only a fraction of its tasks.
Read more →Research & Engineering•7 min read
Our Problems
An overview of the problems we’re working on at Proximal.
Read more →Company•4 min read
Announcing Proximal
We believe data is becoming one of the central research problems in AI, and no one is working on it the right way. Proximal is a research lab for data.
Read more →We're hiring across research and engineering.
If treating data as a first-class research problem sounds like the work you want to be doing, we'd like to hear from you.